This project is a continuation of Project 1B. Over the past decade, this SPORE uncovered a select group of disease-causing gene mutations and conspicuously overexpressed genes. It is now appropriate to select a few of these clues that offer the best promise for rational therapeutic targeting of the molecular differences that distinguish the tumors of individual patients. The specific aims include the characterization of novel cancer mutations, to develop a "pharmacodiagnostic" algorithm for genetically classifying, treating and advising individual patients, and to uncover a molecular basis for overexpressed tumor antigens to accompany antigen-specific therapies now being initiated. Our long-term goals are to develop a new clinical management strategy for patients presenting with invasive pancreatic cancer that will advise the patient through genetic counseling, advise the surgeon and oncologist as to promising therapies, and then provide specific tools for monitoring the dosage and efficacy of targeted therapies. We will utilize tissues form Core 2 to translate discoveries made in this project to early patient screening being developed in proposed projects 1A and 3A and for improved genetic counseling of families in project 3A and in Core 3. Lay summary: In order to benefit pancreatic cancer patients, we will require a deeper understanding of the molecular basis of the disease and the most specific drug targets: to allow risk assessment and detection at the earliest stage, to understand the reasons for its naturally aggressive course, and to devise effective therapy through rational and individualized approaches. This project will help us to understand the key differences among patients that will enable us to select the proper therapies and provide the appropriate genetic counseling.